We’ve pushed several small updates to our Google Analytics Audit tool in the past several days. The number of different types of checks performed by the “Check Health Status” tool has now reached 31 (from 22 a year and a half ago)!
The two most recent changes include checking for individual pages with unrealistically low bounce rate (which is very often a sign for double-tracking on these particular pages) as well as checking for the presence of referrer spam in your traffic source reports. Here are the two changes in more detail:
#1 Unrealistic bounce rate check
We’ve had such a check on sidewide level for quite some time now. However, we were missing cases where on a large site with many different pages and sections only part of the site had double-tracking issues. So, to cope with that we’ve developed page-level checks with a fair guarantee that what we find is reason for concern.
Here is what the report might look like:
Of course, since this can never be a 100% sure check (in some cases a page can have many visits and very very low bounce rate for completely legitimate reasons) it generates a warning only.
#2 Referrer spam detection
Given how prevalent referrer spam is these days, we took a bit long to include referrer spam detection in our tool. This was since we wanted a solution that is scalable so that we can develop it as the spam grows. We now have such a solution and thus from now onwards our Health Status Checker now includes a Referrer Spam Checker in it. The checks are done against our continuously updated referral spam database.
Here is what the report looks like:
Of course, the tool recommends that you use our Auto Spam Filters tool to block referrer spam automatically, however you can always consult our posts on referrer spam for a manual solution. We actually just updated the anti-referral spam filters in there!
Note to Auto Spam Filter users: you will see referrer spam detected if you’ve just recently started using our referrer spam blocker. You might also occasionally see some spam sessions detected even with the blocker in use since no matter how fast we update our database some spam will inevitably slip through before we can react.